BACKGROUND: Sedation is a common and incapacitating clozapine adverse effect, but the factors associated with sedation and its pharmacological management remain poorly studied. METHODS: We conducted a retrospective cohort study based on deidentified electronic clinical records of clozapine-treated patients from the secondary mental health care provider for Cambridgeshire and Peterborough, United Kingdom. We first evaluated cross-sectionally the influence of clozapine dose, clozapine, and norclozapine plasma levels on self-reported hours slept, as a proxy for sedation, using bivariate correlation and then the longitudinal effect of changes in clozapine dose and other 23 medications using linear mixed effect models. We followed 241 clozapine-treated patients for 56 months on average, with 2237 face-to-face assessments in total. RESULTS: Patients slept for a mean of 9.35 h/d, with 46% reporting 10 h/d or more. Cross-sectionally, sleep duration did not correlate with clozapine dose (r = 0.14, P = 0.106), but with clozapine plasma levels (r = 0.38, P < 0.0001) and norclozapine plasma levels (r = 0.25, P = 0.005). Longitudinally, the final mixed-effects model revealed 4 pharmacological variables that had a significant impact on hours slept: clozapine, risperidone augmentation, and atenolol were associated with increased sleep, whereas aripiprazole augmentation was associated with decreased sleep. We found that 20 other psychotropic medications measured were not associated with changes in sleep when added to clozapine. Excess sleep is a clozapine level-dependent adverse effect. CONCLUSIONS: The impact of different augmentation strategies might help clinicians decide on the most adequate strategy, albeit further studies should confirm our results.